[
https://issues.apache.org/jira/browse/HBASE-13376?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=14644890#comment-14644890
]
Ted Yu commented on HBASE-13376:
--------------------------------
Running locally with patch, I saw the following (after test timed out)
{code}
"Thread-4" prio=5 tid=0x00007fc98e8c1800 nid=0x6c13 runnable
[0x0000000116a86000]
java.lang.Thread.State: RUNNABLE
at
org.apache.hadoop.hbase.master.balancer.StochasticLoadBalancer$CostFunction.costFromArray(StochasticLoadBalancer.java:823)
at
org.apache.hadoop.hbase.master.balancer.StochasticLoadBalancer$RegionCountSkewCostFunction.cost(StochasticLoadBalancer.java:950)
at
org.apache.hadoop.hbase.master.balancer.StochasticLoadBalancer.computeCost(StochasticLoadBalancer.java:409)
at
org.apache.hadoop.hbase.master.balancer.StochasticLoadBalancer.balanceCluster(StochasticLoadBalancer.java:276)
- locked <0x0000000751ba3050> (a
org.apache.hadoop.hbase.master.balancer.StochasticLoadBalancer)
at
org.apache.hadoop.hbase.master.balancer.TestStochasticLoadBalancer.testWithCluster(TestStochasticLoadBalancer.java:662)
at
org.apache.hadoop.hbase.master.balancer.TestStochasticLoadBalancer.testWithCluster(TestStochasticLoadBalancer.java:651)
at
org.apache.hadoop.hbase.master.balancer.TestStochasticLoadBalancer.testMidCluster2(TestStochasticLoadBalancer.java:478)
{code}
> Improvements to Stochastic load balancer
> ----------------------------------------
>
> Key: HBASE-13376
> URL: https://issues.apache.org/jira/browse/HBASE-13376
> Project: HBase
> Issue Type: Improvement
> Components: Balancer
> Affects Versions: 1.0.0, 0.98.12
> Reporter: Vandana Ayyalasomayajula
> Assignee: Vandana Ayyalasomayajula
> Priority: Minor
> Attachments: 13376-v2.txt, HBASE-13376.patch, HBASE-13376_0.98.txt,
> HBASE-13376_0.txt, HBASE-13376_1.txt, HBASE-13376_1_1.txt,
> HBASE-13376_2_branch-1.patch, HBASE-13376_98.patch, HBASE-13376_branch-1.patch
>
>
> There are two things this jira tries to address:
> 1. The locality picker in the stochastic balancer does not pick regions with
> least locality as candidates for swap/move. So when any user configures
> locality cost in the configs, the balancer does not always seems to move
> regions with bad locality.
> 2. When a cluster has equal number of loaded regions, it always picks the
> first one. It should pick a random region on one of the equally loaded
> servers. This improves a chance of finding a good candidate, when load picker
> is invoked several times.
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
